INDEX

Explanations

words related to directional trends, specifically looking for words indicating an upward direction

oai_token-act-pair · gpt-3.5-turbo

phrases describing trends in data, particularly upward or downward movements

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.7.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

DN

-0.79

BuyableInstoreAndOnline

-0.74

abase

-0.73

TOP

-0.71

 Mamm

-0.70

mong

-0.68

tek

-0.67

PLIED

-0.66

rol

-0.65

pty

-0.65

POSITIVE LOGITS

 trajectory

0.98

 traject

0.96

 pressure

0.87

 revisions

0.85

stairs

0.82

volent

0.81

 spiral

0.81

grading

0.78

pressure

0.77

srf

0.76

Activations Density 0.027%

words related to directional trends, specifically looking for words indicating an upward direction

phrases describing trends in data, particularly upward or downward movements

No Comments

No Known Activations