INDEX

Explanations

proper nouns that potentially represent locations, names, or specific entities

oai_token-act-pair · gpt-3.5-turbo

the word "une" in various contexts

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

loo

-0.89

ivation

-0.81

ories

-0.80

pread

-0.78

İĭ

-0.75

rieg

-0.74

Ö¼

-0.72

draw

-0.72

ngth

-0.72

orial

-0.71

POSITIVE LOGITS

arthed

1.02

quist

0.67

cker

0.65

nels

0.64

lected

0.64

Mik

0.62

hill

0.62

 Marble

0.60

une

0.59

uter

0.58

Activations Density 0.016%

proper nouns that potentially represent locations, names, or specific entities

the word "une" in various contexts

No Comments

No Known Activations