INDEX

Explanations

legal case-related terms

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 @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

erd

-0.88

atal

-0.86

odes

-0.86

pez

-0.84

oded

-0.82

atomic

-0.81

attery

-0.80

undai

-0.77

olen

-0.76

emouth

-0.75

POSITIVE LOGITS

 filed

0.87

cases

0.86

making

0.80

suit

0.78

 against

0.77

suits

0.73

 involving

0.72

 Heller

0.72

Attach

0.70

Ruk

0.68

Activations Density 0.033%

legal case-related terms

references to legal cases

No Comments

No Known Activations