INDEX

Explanations

words related to financial institutions, especially banks

oai_token-act-pair · gpt-3.5-turbo

mentions of banks

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

 Loving

-0.74

Rust

-0.69

witz

-0.67

 Hawth

-0.67

imens

-0.66

reads

-0.66

 Ging

-0.66

Hunt

-0.66

 Warhammer

-0.65

 Nato

-0.62

POSITIVE LOGITS

rupt

1.08

rolled

1.06

notes

0.99

roll

0.98

note

0.97

rollers

0.93

roller

0.87

gage

0.87

 banks

0.87

sters

0.87

Activations Density 0.036%

words related to financial institutions, especially banks

mentions of banks

No Comments

No Known Activations