INDEX

Explanations

words related to health and well-being

oai_token-act-pair · gpt-3.5-turbo

terms related to beneficial or detrimental effects

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

buck

-0.82

oute

-0.72

pler

-0.64

metal

-0.63

ascus

-0.62

hyde

-0.62

rine

-0.61

bish

-0.61

herer

-0.60

Tube

-0.60

POSITIVE LOGITS

 outweigh

0.77

icial

0.77

 outcomes

0.73

 effects

0.72

 synerg

0.72

 influence

0.71

nerg

0.70

 influences

0.69

 effect

0.68

 Advantage

0.68

Activations Density 0.079%

words related to health and well-being

terms related to beneficial or detrimental effects

No Comments

No Known Activations