INDEX

Explanations

proper nouns, specifically names containing the sequence "imon"

oai_token-act-pair · gpt-3.5-turbo

names or terms that reference individuals or specific associated entities

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

LEY

-0.74

IGH

-0.73

ees

-0.73

erer

-0.72

rees

-0.72

ENC

-0.71

REE

-0.67

giving

-0.67

hani

-0.65

UGE

-0.65

POSITIVE LOGITS

ãĥĨãĤ£

0.87

ious

0.82

ium

0.79

ial

0.76

omon

0.76

otor

0.76

ned

0.75

stration

0.71

ihilation

0.70

ica

0.70

Activations Density 0.029%

proper nouns, specifically names containing the sequence "imon"

names or terms that reference individuals or specific associated entities

No Comments

No Known Activations