INDEX

Explanations

political and societal terms or descriptions

oai_token-act-pair · gpt-3.5-turbo

constructions that emphasize associations or descriptors of groups

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

ãĥĦ

-0.74

jen

-0.71

atre

-0.69

raft

-0.67

hene

-0.66

itarian

-0.66

Tomorrow

-0.66

thens

-0.65

riks

-0.65

sburgh

-0.64

POSITIVE LOGITS

 whereby

1.01

 which

0.99

 implying

0.95

 namely

0.91

 Lastly

0.90

 suggesting

0.90

 wherein

0.89

which

0.89

 allowing

0.88

 lest

0.87

Activations Density 0.507%

political and societal terms or descriptions

constructions that emphasize associations or descriptors of groups

No Comments

No Known Activations