INDEX

Explanations

mentions of allies or allies in political contexts

oai_token-act-pair · gpt-3.5-turbo

instances of the word "ally" and related terms

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ammers

-0.86

usage

-0.85

cale

-0.85

ummer

-0.84

essee

-0.80

umbers

-0.80

pace

-0.78

hift

-0.78

omsday

-0.78

uits

-0.78

POSITIVE LOGITS

 ally

0.92

 Ally

0.87

="#

0.74

 colonel

0.72

=]

0.68

Vic

0.67

Fei

0.67

 Allies

0.67

 confid

0.67

 comrade

0.66

Activations Density 0.017%

mentions of allies or allies in political contexts

instances of the word "ally" and related terms

No Comments

No Known Activations