INDEX

Explanations

averages or statistics related to performance

oai_token-act-pair · gpt-3.5-turbo

statistical measurements related to averages in performance metrics

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

aina

-0.65

ighter

-0.65

lean

-0.63

nuts

-0.62

nda

-0.62

hair

-0.61

nut

-0.61

 Syndrome

-0.61

nee

-0.60

 syndrome

-0.60

POSITIVE LOGITS

imates

0.89

oday

0.82

 averages

0.81

uates

0.76

 approx

0.75

uate

0.74

imum

0.74

imate

0.73

nesota

0.72

icals

0.72

Activations Density 0.019%

averages or statistics related to performance

statistical measurements related to averages in performance metrics

No Comments

No Known Activations