INDEX

Explanations

words related to actions and events

oai_token-act-pair · gpt-3.5-turbo

instances of actions or experiences involving social interaction and personal perception

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

 Stability

-0.71

insula

-0.67

 feasibility

-0.67

tein

-0.64

 integration

-0.64

 convergence

-0.62

 Combine

-0.61

LF

-0.60

 shader

-0.60

ces

-0.57

POSITIVE LOGITS

 aback

0.95

oing

0.80

ĸļ

0.80

inged

0.77

ãĤ¼

0.74

ptin

0.74

©¶æ¥µ

0.74

zeb

0.72

anked

0.71

chy

0.71

Activations Density 0.293%

words related to actions and events

instances of actions or experiences involving social interaction and personal perception

No Comments

No Known Activations