INDEX

Explanations

references to locations where incidents or events have occurred

oai_token-act-pair · gpt-3.5-turbo

references to a specific location in news reports of incidents or events

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

thood

-0.92

 Seym

-0.81

raltar

-0.74

atever

-0.71

glomer

-0.71

haps

-0.68

iago

-0.68

anson

-0.67

dinand

-0.67

yip

-0.66

POSITIVE LOGITS

EMS

0.79

 scene

0.78

Dialogue

0.78

 Scene

0.77

 Reconstruction

0.77

ewitness

0.76

scene

0.73

Capture

0.73

âĶĢ

0.72

 unfolding

0.68

Activations Density 0.022%

references to locations where incidents or events have occurred

references to a specific location in news reports of incidents or events

No Comments

No Known Activations