INDEX

Explanations

phrases related to whether things meet expectations or not

oai_token-act-pair · gpt-3.5-turbo

phrases that indicate expectations or anticipated outcomes

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

agra

-0.77

nan

-0.74

tex

-0.73

vertisement

-0.73

aina

-0.72

tein

-0.72

aska

-0.70

stein

-0.70

tha

-0.69

anski

-0.68

POSITIVE LOGITS

LY

0.74

ICAL

0.70

 Pascal

0.67

ĭ

0.65

OLOG

0.62

 curve

0.61

 unexpected

0.61

antly

0.61

 consequence

0.61

 plur

0.60

Activations Density 0.031%

phrases related to whether things meet expectations or not

phrases that indicate expectations or anticipated outcomes

No Comments

No Known Activations