INDEX

Explanations

references to hidden or confidential information

oai_token-act-pair · gpt-3.5-turbo

references to the concept of secrets

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

annis

-0.85

gaard

-0.84

lihood

-0.80

ulf

-0.79

odcast

-0.78

kins

-0.78

older

-0.76

ucc

-0.76

thood

-0.75

ð

-0.75

POSITIVE LOGITS

arial

1.19

ariat

1.18

 ingredient

0.93

 sauce

0.84

rets

0.81

 secret

0.81

 handshake

0.81

 ballot

0.80

 stash

0.78

 unfocusedRange

0.77

Activations Density 0.027%

references to hidden or confidential information

references to the concept of secrets

No Comments

No Known Activations