INDEX

Explanations

references to legal cases or judicial proceedings

oai_token-act-pair · gpt-3.5-turbo

references to legal cases

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 livest

-0.77

erd

-0.71

arus

-0.67

odes

-0.67

eton

-0.66

lasses

-0.66

icc

-0.66

amaru

-0.66

ode

-0.65

atomic

-0.65

POSITIVE LOGITS

 against

1.04

 involving

1.03

 prosecuted

0.97

 filed

0.91

 alleging

0.89

 proceedings

0.85

against

0.83

 pending

0.81

 litigation

0.81

 courthouse

0.80

Activations Density 0.053%

references to legal cases or judicial proceedings

references to legal cases

No Comments

No Known Activations