INDEX

Explanations

numbers that are below a certain threshold

oai_token-act-pair · gpt-3.5-turbo

references to measurements or values that are below specific thresholds

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

RAW

-0.87

Sense

-0.75

Language

-0.72

Mail

-0.71

Creat

-0.70

=-=-=-=-=-=-=-=-

-0.69

MM

-0.69

 Instruments

-0.69

Frames

-0.69

ãĤ£

-0.68

POSITIVE LOGITS

ground

0.85

cedented

0.75

lip

0.74

 grade

0.74

 hypers

0.74

pping

0.72

uckland

0.71

grade

0.71

sighted

0.71

raviolet

0.70

Activations Density 0.030%

numbers that are below a certain threshold

references to measurements or values that are below specific thresholds

No Comments

No Known Activations