INDEX

Explanations

names containing the letters "mun"

oai_token-act-pair · gpt-3.5-turbo

the mention of "mun" frequently, indicating a focus on a specific entity or term associated with authority or governance

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 heels

-0.73

 tumult

-0.64

 markup

-0.62

 rave

-0.62

 Shining

-0.62

 vodka

-0.62

 follower

-0.61

 Rise

-0.60

 Price

-0.60

 dashed

-0.60

POSITIVE LOGITS

mun

5.08

munition

1.52

mun

1.40

mu

1.12

mud

1.11

unicip

1.08

mon

1.04

mos

1.02

bara

0.98

mond

0.98

Activations Density 0.012%

names containing the letters "mun"

the mention of "mun" frequently, indicating a focus on a specific entity or term associated with authority or governance

No Comments

No Known Activations