INDEX

Explanations

references to seismic events by their magnitude

oai_token-act-pair · gpt-3.5-turbo

references to earthquake magnitudes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

bones

-0.81

ohydrate

-0.74

 Byrne

-0.73

starter

-0.73

foundation

-0.73

comes

-0.72

nee

-0.67

 Judges

-0.65

stairs

-0.64

horse

-0.64

POSITIVE LOGITS

 magnitude

1.19

 scale

0.91

 equival

0.85

 inacc

0.82

externalActionCode

0.79

ricane

0.78

 impact

0.76

 amplitude

0.74

 amount

0.74

 sensitivity

0.73

Activations Density 0.006%

references to seismic events by their magnitude

references to earthquake magnitudes

No Comments

No Known Activations