INDEX

Explanations

references to specific fiscal years

oai_token-act-pair · gpt-3.5-turbo

Triggered by @bot

terms related to fiscal policy and government budgets

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 12-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.hook_resid_post

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.11.hook_resid_post

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

Companies

-0.81

ipedia

-0.72

xual

-0.70

hra

-0.69

gom

-0.68

Pie

-0.65

esta

-0.65

ById

-0.64

href

-0.64

YL

-0.64

POSITIVE LOGITS

 cliff

1.09

 consolidation

0.93

 stimulus

0.91

 prud

0.85

 restraint

0.85

 year

0.85

 responsibility

0.82

 discipline

0.79

 cliffs

0.79

 Cliff

0.78

Activations Density 0.032%

references to specific fiscal years

terms related to fiscal policy and government budgets

No Known Activations