INDEX

Explanations

numerical values associated with software releases or technical specifications

oai_token-act-pair · gpt-3.5-turbo

numerical data and values in a technical context

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

erity

-0.84

lehem

-0.72

agna

-0.70

netflix

-0.70

Lago

-0.67

UAL

-0.63

olson

-0.63

Frames

-0.62

pict

-0.61

 conduc

-0.60

POSITIVE LOGITS

xff

1.07

resents

0.89

0.77

ãĥ³

0.74

0.73

0.70

oldemort

0.70

reys

0.69

MQ

0.68

Activations Density 0.024%

numerical values associated with software releases or technical specifications

numerical data and values in a technical context

No Comments

No Known Activations