INDEX

Explanations

terms related to a specific district, potentially linked to legal or administrative matters

oai_token-act-pair · gpt-3.5-turbo

references to specific districts, particularly districts 8 and 9

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ipedia

-0.81

ocaust

-0.74

odus

-0.73

glers

-0.72

obo

-0.72

fuel

-0.72

racuse

-0.71

ollo

-0.68

pez

-0.68

obs

-0.67

POSITIVE LOGITS

rict

0.94

 superintendent

0.92

wide

0.86

 boundaries

0.85

 attorney

0.83

 Superintendent

0.80

 district

0.78

 magistrate

0.77

 districts

0.76

ricting

0.76

Activations Density 0.027%

terms related to a specific district, potentially linked to legal or administrative matters

references to specific districts, particularly districts 8 and 9

No Comments

No Known Activations