INDEX

Explanations

places or locations

oai_token-act-pair · gpt-3.5-turbo

descriptive phrases about locations or characteristics of places

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

uctor

-0.78

ctive

-0.74

 warranties

-0.73

duration

-0.71

illon

-0.67

argo

-0.65

 transactions

-0.65

enary

-0.64

ideos

-0.64

sers

-0.64

POSITIVE LOGITS

 situated

1.15

 home

1.15

 located

1.07

 populated

1.06

 bustling

1.05

 frequ

1.04

 dotted

1.04

 littered

1.03

 inhabited

1.00

 densely

0.98

Activations Density 0.226%

places or locations

descriptive phrases about locations or characteristics of places

No Comments

No Known Activations