INDEX

Explanations

words related to directional movements or trends like "upward" or "downward"

oai_token-act-pair · gpt-3.5-turbo

terms related to upward and downward trends

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

abase

-0.77

mong

-0.76

DN

-0.73

 Mamm

-0.73

TOP

-0.70

BuyableInstoreAndOnline

-0.70

ument

-0.66

mberg

-0.65

tek

-0.65

ingo

-0.64

POSITIVE LOGITS

 traject

0.94

 trajectory

0.93

¥ŀ

0.93

stairs

0.88

comings

0.85

graded

0.85

srf

0.84

grading

0.83

 slopes

0.80

 spiral

0.79

Activations Density 0.020%

words related to directional movements or trends like "upward" or "downward"

terms related to upward and downward trends

No Comments

No Known Activations