INDEX

Explanations

words related to negative impact or harm

oai_token-act-pair · gpt-3.5-turbo

terms related to negative impacts 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 @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

-0.79

ariat

-0.78

rows

-0.71

illon

-0.70

Kings

-0.68

ebus

-0.65

peak

-0.64

hung

-0.64

onne

-0.64

pper

-0.64

POSITIVE LOGITS

 undermin

0.86

 impact

0.75

 effects

0.74

ãĥ¼ãĥĨ

0.73

 synerg

0.73

 aspects

0.73

 harmful

0.73

 alien

0.72

 disrupt

0.71

 harm

0.71

Activations Density 0.047%

words related to negative impact or harm

terms related to negative impacts or detrimental effects

No Comments

No Known Activations