INDEX

Explanations

words related to reasons or intentions

oai_token-act-pair · gpt-3.5-turbo

references to motives or reasons behind actions

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

ummer

-0.98

 Leopard

-0.80

enegger

-0.80

semble

-0.78

iannopoulos

-0.76

thumbnails

-0.73

opy

-0.71

rooms

-0.71

scl

-0.68

room

-0.66

POSITIVE LOGITS

 motives

1.06

 motive

0.99

 motivations

0.95

 motivation

0.88

 justifying

0.87

 behind

0.82

 cele

0.82

TextColor

0.82

 motivated

0.79

 reasoning

0.75

Activations Density 0.020%

words related to reasons or intentions

references to motives or reasons behind actions

No Comments

No Known Activations