INDEX

Explanations

proper nouns appearing in the text

oai_token-act-pair · gpt-3.5-turbo

mentions of a specific name or title that begins with "Der."

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

hetti

-0.83

INTER

-0.71

 Dragonbound

-0.70

[|

-0.67

sburgh

-0.67

 Samoa

-0.66

odcast

-0.65

ships

-0.64

margin

-0.63

hews

-0.63

POSITIVE LOGITS

bys

0.98

ricks

0.93

bil

0.90

 Spiegel

0.88

iving

0.88

mal

0.88

rek

0.87

isively

0.86

ived

0.84

ivation

0.84

Activations Density 0.009%

proper nouns appearing in the text

mentions of a specific name or title that begins with "Der."

No Comments

No Known Activations