INDEX

Explanations

phrases related to small measurements or amounts

oai_token-act-pair · gpt-3.5-turbo

references to proportions and fractions

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

OWS

-0.76

TOP

-0.70

ãĤ¹ãĥĪ

-0.68

ãĤ¤ãĥĪ

-0.68

Reviewer

-0.67

Ãį

-0.66

Trend

-0.66

ennes

-0.65

uments

-0.64

çİĭ

-0.64

POSITIVE LOGITS

dozen

0.98

 dozen

0.97

icum

0.89

(~

0.82

omething

0.82

(<

0.81

nown

0.75

abyte

0.75

 tolerated

0.74

 kilomet

0.72

Activations Density 0.433%

phrases related to small measurements or amounts

references to proportions and fractions

No Comments

No Known Activations