INDEX

Explanations

italicized words

oai_token-act-pair · gpt-3.5-turbo

punctuation and structural elements within sentences

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

enhagen

-0.73

INTON

-0.69

 scrut

-0.69

quickShipAvailable

-0.67

AMP

-0.64

emort

-0.63

 Accountability

-0.62

 unintention

-0.61

 cryst

-0.61

TPS

-0.61

POSITIVE LOGITS

sic

1.03

ensis

0.89

tnc

0.77

dust

0.74

edu

0.70

mun

0.69

literally

0.68

oult

0.68

rosis

0.68

utils

0.67

Activations Density 0.871%

italicized words

punctuation and structural elements within sentences

No Comments

No Known Activations