INDEX

Explanations

generic terms or labels

oai_token-act-pair · gpt-3.5-turbo

references to generic or generalized concepts and terms

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

oir

-0.77

kers

-0.76

otos

-0.76

rey

-0.74

Jenn

-0.73

cano

-0.73

 Ferry

-0.71

orah

-0.68

alos

-0.67

izoph

-0.67

POSITIVE LOGITS

ization

0.83

 applic

0.79

 generic

0.79

 differentiation

0.75

ALLY

0.74

ature

0.74

isable

0.72

isation

0.72

 interchange

0.71

 error

0.70

Activations Density 0.010%

generic terms or labels

references to generic or generalized concepts and terms

No Comments

No Known Activations