INDEX

Explanations

words related to geographic locations, particularly straits

oai_token-act-pair · gpt-3.5-turbo

mentions of geographical locations, specifically straits

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

kus

-0.76

MAT

-0.72

SEC

-0.71

ERAL

-0.70

PRES

-0.68

âķĲâķĲ

-0.68

MN

-0.67

Miller

-0.67

GRE

-0.64

erg

-0.64

POSITIVE LOGITS

 Strait

1.41

rait

0.96

itudinal

0.79

 Islanders

0.79

raits

0.79

 stra

0.77

 separating

0.75

 Island

0.74

 sands

0.72

istani

0.71

Activations Density 0.010%

words related to geographic locations, particularly straits

mentions of geographical locations, specifically straits

No Comments

No Known Activations