INDEX

Explanations

references to a specific location named "Grove."

oai_token-act-pair · gpt-3.5-turbo

references to the term "gro" and its variations, likely related to grocery or grocery-related contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

DSL

-0.72

 Mavericks

-0.67

CPC

-0.67

icts

-0.65

 gulf

-0.65

è¦ļéĨĴ

-0.64

awed

-0.61

icated

-0.61

 Falling

-0.61

 Extreme

-0.60

POSITIVE LOGITS

gro

1.19

tesque

1.16

veyard

0.95

 Grove

0.92

lett

0.85

ovy

0.84

tto

0.79

lore

0.77

=#

0.77

llo

0.74

Activations Density 0.007%

references to a specific location named "Grove."

references to the term "gro" and its variations, likely related to grocery or grocery-related contexts

No Comments

No Known Activations