INDEX

Explanations

words expressing the idea of adherence or compliance with something

oai_token-act-pair · gpt-3.5-turbo

instances of the word "follow" and its variations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

vity

-0.79

bodied

-0.77

voy

-0.71

cer

-0.69

cery

-0.66

rait

-0.66

atonin

-0.65

Ñı

-0.65

ividual

-0.65

ancing

-0.64

POSITIVE LOGITS

 suit

1.44

 directions

0.98

 closely

0.93

pend

0.87

 instructions

0.87

 protocol

0.83

suit

0.82

 Suit

0.82

 strict

0.81

 blindly

0.78

Activations Density 0.044%

words expressing the idea of adherence or compliance with something

instances of the word "follow" and its variations

No Comments

No Known Activations